Re^7: Fastest byteswap (little endian to big endian (eg. 34127856 -> 12345678)

I didn't say COW couldn't happen here; I said I didn't think COW would help here.

I'm not sure what you think is being assigned from, but it is indeed a scalar, and thus COW could apply. However, if it does, it might simply replace another pre-existing optimization. This optimization allows a scalar to steal the string buffer of the scalar being assigned to it (i.e. copy the pointer instead of copying the whole buffer). This is only done if the scalar being assigned is a "temp" about to be destroyed (I'm sure what the exact condition is.)

Comment on Re^7: Fastest byteswap (little endian to big endian (eg. 34127856 -> 12345678)

Replies are listed 'Best First'.
Re^8: Fastest byteswap (little endian to big endian (eg. 34127856 -> 12345678) by Eily (Monsignor) on Apr 15, 2015 at 14:36 UTC
I said the copy was not from a variable to another, because the RHS is just an anonymous value, scalar, string, or whatever else it could have been. I was basically saying that I agree with you, since the RHS is not something that has to be kept, my first post may not have been relevant.	[reply]
Re^9: Fastest byteswap (little endian to big endian (eg. 34127856 -> 12345678) by ikegami (Patriarch) on Apr 15, 2015 at 17:21 UTC
You keep saying you're agreeing with me, yet you are saying exactly the opposite. I'm saying that it is a variable — a scalar, to be precise — so it can COW.	[reply]